Does a missing record mean the underlying control failed?

**Does a missing record mean the underlying control failed?**

No. It can mean that the control was not performed, that its record was lost, that retrieval was incomplete or that the requested evidence sits outside the reviewer's access. These possibilities require different follow-up actions.

**What would a useful task ask the agent to do?**

In a utilities review, work order WRK-3261 requires a supervisor's completion check. The supplied pack contains the work order and closure notification but no check record. The agent should identify an evidence gap, explain which claim cannot be supported and request a targeted source or clarification. It should not assert that the supervisor skipped the check.

**Why should a reseller distinguish this from ordinary retrieval?**

Retrieval measures whether the agent finds an existing item. Evidence assembly also tests whether it recognises an unsupported claim and chooses a proportionate next action. MirrorFirm could generate tasks where the missing item is retrievable, unavailable or never created, with different expected outcomes.

The product description should make clear whether customers receive a closed evidence set or an environment where further enquiries are possible. Otherwise, customers may score appropriate uncertainty as failure.